In this episode of @HPCpodcast, Shahin and Doug caught up with Steve Conway, former senior advisor at Hyperion Research and now head of his own consulting firm, Conway Communications. A well-known HPC executive and analyst who was formerly with IDC and Cray, Conway engages us in a wide ranging discussion – we start with edge HPC and trends towards massively-distributed, massively-heterogeneous computing, which takes us to the convergence of HPC and AI, mixed precision spectrum, the importance of simulation, the impact of exascale on general computing, the “indigenous technology” trend that has taken root in the U.S., China and Europe and it’s impact on scientific collaboration, U.S.-China funding model differences, and the necessary ingredients for attracting top talent.
@HPCpodcast: AWS, Sun, eBay, Netflix (and Others) Vet Adrian Cockcroft Talks Cloud HPC-AI and the Amazon Sustainability Data Initiative
In our converesation with Adrian Cockcroft, we start with Netflix’s move to the cloud, a significant event that helped put cloud computing on the map. Then it’s on to Environment, Sustainability, and Governance (ESG), Formula-1 racing, and cloud configurations and interconnects for HPC and AI workloads. And we talk about accessing the Amazon Sustainability Data Initiative (ASDI) and its petabytes of data, including weather observations, ocean temperatures, climate projection data and satellite imagery, used by researchers and scientists….
@HPCpodcast: A Look at HPC Marketing Strategies with Mike Bernhardt — It’s More Than Speeds-and-Feeds
Because HPC is barrier-pushing technology it’s understandable that HPC marketers often focus their messaging on speeds-and-feeds. But Mike Bernhardt, one of the most recognized names in supercomputing strategic marketing and communications, argues that an over-emphasis on this can be a big, if natural, mistake. Yes, speeds-and-feeds is important, but there are other critical aspects of corporate and product positioning that also need attention if organizations are to attract and retain not only customers but also employees, investors, business partners and the good opinion of government regulators, among other audiences.
@HPCpodcast: Horst Simon on DOE’s Post-Exascale HPC Vision – More Flexibility, More Vendor Diversity; the Start of a New Leadership-class Supercomputing Market?
Following DOE’s next-gen supercomputing Request for Information (RFI) issued last week, we discussed what it all may mean with Dr. Horst Simon, Special Advisor to the Laboratory Director at Lawrence Berkeley National Laboratory and co-editor of the TOP500 list since 2000. He takes us through the current (though fungible) state of DOE’s post-exascale vision, the implications of an Advanced Computing Ecosystem (ACE) outlined in the RFI and the possible emergence of a new, more vendor-diverse leadership-class systems market.
@HPCpodcast: At the HPC User Forum — A Frontier Visit, the State of Quantum and Supercomputing Center Staffing Woes
In this episode of the @HPCpodcast we focus on the HPC User Forum, run by industry analyst firm Hyperion Research and held last week at Oak Ridge National Laboratory in Tennessee. The conference included an opportunity to get a viewing of Frontier, newly crowned as the world’s most powerful computer and the first to break the exascale barrier. You may have heard about it. The User Forum covered an array of topics, including the drama involved in getting Frontier over the exascale finish line in time for the mid-year TOP500 list. Also discussed was what comes next for leadership-class supercomputing, the state of quantum computing, and staffing problems at HPC sites that train graduates for careers in HPC and AI and who then jump to big tech companies for more money.
@HPCpodcast: Google’s Lifelike LaMDA AI Chatbot and Questions of Being or Nothingness
When a tech news story gets talked about on sports radio, you know it’s gone very viral. That’s what happened last week with the story about a Google engineer, Blake Lemoine, who declared that the company’s AI chatbot, LaMDA, is a person with rights. Lemoine promptly got suspended by Google for his trouble, and he says he won’t be surprised if he gets fired. In this episode of the @HPCpodcast, Shahin Khan of OrionX.net and insideHPC editor-in-chief Doug Black talk about LaMDA’s amazingly lifelike conversational capability, how it can ingest books and research papers and share insights about them in real time (i.e., during conversations), deep fake-related ethical questions raised by LaMDA, the urgency of thoughtful social policies based on ethical and legal frameworks and philosophical issues of sentience, being and nothingness – artificial and otherwise.
@HPCpodcast: On the Scene at ISC 2022 – HPE, AMD Make TOP500 News; Intel Makes News of Its Own
ISC 2022 in Hamburg was notable for a number of reasons – it was not only the first in-person ISC since 2019, it also provided a plethora of major news. This included big changes at the top of the TOP500 list of the world’s most powerful supercomputers, and the Frontier HPC system at Oak Ridge National Lab surpassing of the exascale milestone. While AMD, whose chips power Frontier, and HPE, which built Frontier, were the conference’s spotlight vendors, Intel also made some impressive product announcements, as analyzed in this discussion by Shahin Khan. You can find our podcasts at insideHPC’s @HPCpodcast page, on Twitter and at the OrionX.net blog. Here’s the RSS feed.
@HPCpodcast: Jack Dongarra Talks Turing Award, the TOP500 and the Past and Future of Supercomputing
At @HPCpodcast we’ve been fortunate to host some highly distinguished computer scientists and HPC thinkers, people who have shaped supercomputing as the technology has advanced human knowledge and taken on the world’s most vexing challenges. Today we welcome Jack Dongarra who was recently honored with the ACM Turing Award for “Pioneering Concepts and Methods Which Resulted in World-Changing Computations.” As the ACM said in its award announcement, “Dongarra has led the world of high-performance computing through his contributions to….
@HPCpodcast: Exascale in China and a Philosophical Turn on Riding Advanced Tech to Super-wealth and Media Power
The mystery shrouded in a riddle that is the state of exascale supercomputing in China is the main topic of this week’s @HPCpodcast episode. Setting off a re-focus on the recurring topic: the release of a research paper from PRC university scientists on the use of the Sunway TaihuLight system, the no. 4-ranked supercomputer on the Top500 list, in “quantum many-body problems,” which are problems of extreme complexity and scale.
@HPCpodcast: HPC Storage Rock Star Gary Grider Talks How We Got Here and Where We’re Going
http://orionx.net/wp-content/uploads/2022/04/021@HPCpodcas_Storage-with-Gary-Grider_20220420.mp3 Shahin and I are in violent agreement: if you’re interested in high performance storage, if major milestone in the development of HPC storage technology over the last 30-plus years interests you, if you want a peek at future forward leaps in HPC storage technology, then this is an @HPCpodcast episode for you. Our special […]
